# -.-|m { input_fold: hide }
%load_ext pretty_jupyter
Introduction¶
For this assignment I decided using the 🔬Melanoma Cancer Image Dataset. It consists of 13 900 high-resolution images in test and train sets, both divided in two categorical fodlers - Benign and Malignant. The test folders comprise ot 1 000 photos each.
For the choice of the pre-trained model I searched for previous uses and researches. Based on the performance comparison in the Sagar (2020) research I decided to opt for the ResNet50V2 model.
a = 10
b = 5